J. Am. Chem. SOC.1981, 103, 5965-5967 that in the biosynthesis of the growing polyhydroxy fatty acid chain, P-ketoacyl CoA reduction can occur with either stereochemistry. For the moment no information is available as to how the secondary methyl stereochemistry at (2-2, C-4, C-6, and C-24 is set.26 While further experiments, now in progress, will be required to determine directly the origins of the three remaining ether oxygens as well as that of the C-26 hydroxyl, it is likely that all four atoms are derived from molecular oxygen. Extending the original suggestion of Westley,’ it is therefore interesting to speculate that the first formed polyfunctional fatty acid would be the all-(!?)-triene 627which could undergo epoxidation to give the (12R,13R,16R,17R,20S,21S)-triepoxide 7 (Scheme 111). Attack of the C-5 hydroxyl of 7 at the C-9 carbonyl carbon would initiate a cascade of ring closures to generate all five ether rings of monensin. These results are also in accord with similar findings by Hutchinson et al. on the biosynthesis of the aromatic polyether lasalocid, reported in the accompanying communication^.^^
5965
generated from PhHgCC12Br to give the insertion product in high yields2 In connection with the current interest in oxyanionic substituent effects,’ we have examined the reactivity of carbenes toward alkoxide anions. Herein we wish to report a novel effect of oxyanion substituent which greatly facilitates the insertion of (phenylthio)carbene4 into the a-C-H bond of alkoxide anion^.^ To a mixed suspension of sodium allylalkoxide (la, 2.5 equiv) and t-BuOK (2.5 equiv) in T H F was added a T H F solution of chloromethyl phenyl sulfide, and the total mixture was stirred for 0.5 h at 0 “C. After aqueous workup followed by column chromatography, l-(phenylthi0)-3-buten-2-01 (2a) and allyl (pheny1thio)methyl ether (3a) were isolated in 29% and 64% yield, respectively (eq 1). The formal insertion product 2a could be I
H
la
Acknowledgment. This work was supported by NIH Grant GM 22172. Strains of S. cinnamonensis as well as samples of monensin were generous gifts of the Eli Lilly Co. The [I3C]cyanide was obtained from the National Stable Isotopes Resource, Los Alamos Scientific Laboratory, supported by the US.DOE and the N I H (RR-00962). The Bruker WM 250 N M R spectrometer used in this study was purchased with funds provided by the N S F and the Montedison Group of Milan. We would like to thank Professor Richard Hutchinson of the University of Wisconsin for frequent mutual exchanges of information and for generously offering to delay publication of his own in order to allow simultaneous submission of our manuscripts. (26) The methyl group stereochemistry at C-18 and (2-22 is presumably determined as a consequence of the reduction of cu-methyl, a,D-unsaturatedacyl-CoA intermediates, analogous to those of normal fatty acid biosynthesis: F. Lynen, Biochem. J., 102, 381 (1967); K. Bloch and D. Vance, Annu. Rev. Eiochem., 46,263 (1977); B. Sedgwick and J. W. Cornforth, Eur. J. Eiochem., 75,465 (1977); B. Sedgwick, J. W. Cornforth, S. J. French, R. T. Gray, E. Kelstrup, and P. Willadsen, ibid., 75, 481 (1977); B. Sedgwick, C. Morris, and S. J . French, J . Chem. SOC.,Chem. Commun., 193 (1978). (27) Each (E)-olefin could be formed by either syn dehydration of an erythro-a-alkyl-0-hydroxyacyl-CoA or anti dehydration of the corresponding fhreo-a-alkyl-j3-hydroxyacyl-CoA.In the latter regard we note the threo relationship of the C-2,-3, C-4,-5, and C-6,-7 centers of monensin itself. The same threo stereochemistry is prevalent in all known macrolides, as summarized by Celmer’s rules.** (28) W. D. Celmer, Pure Appl. Chem., 28,413 (1971); W. D. Celmer, J . Am. Chem. SOC.,87, 1801 (1965). (29) C. R. Hutchinson, M. M. Sherman, J. C. Vederas, and T. T. Nakashima, J . Am. Chem. SOC.,first paper in this series; C. R. Hutchinson, M. M. Sherman, J. C. Vederas, A. G . McInnes, and J . A. Walters, ibid., second paper in this series.
Reaction of Alkoxides with (Pheny1thio)carbene: A Novel Oxyanionic Substituent Effect on the C-H Insertion of Carbenes Toshiro Harada and Akira Oku* Department of Chemistry, Kyoto Institute of Technology Matsugasaki, Sakyo-ku, Kyoto 606, Japan Received May 20, 1981 While the addition of carbenes to alkenes has been utilized as an efficient tool for C-C bond formation in organic synthesis, the insertion of carbenes has been overlooked owing to its less selective nature.’ An activation of a particular C-H bond seems to be a requisite for the selective insertion to occur. For example, Seyferth et al. have reported that the a-C-H bond of tetrahydrofuran shows enhanced reactivity toward dichlorocarbene (1) (a) Kirmse, W. “Carbene Chemistry”, 2nd ed.; Academic Press: New York, 1971. (b) Parham, E. P.; Schweizer, E. E. Org. Reacf. 1963, 13, 55. (c) Burke, S. D.; Grieco, P. A. Ibid. 1971, 26, 361. (d) Simmons, H. E.; Cairns, T. L.; Vladuchick, S. A,; Hoiness, C. M. Ibid. 1973, 20, 1.
2a
(1)
-OCH,SPh
3a
produced via a Wittig rearrangement of the first formed ether 3a as shown in eq 2. However, the possibility is ruled out as 3a
I-BuOK/THF reflux
-
~
O
C
H
Z
S
+ P ~ 2a
4
(2)
L w p r
follows: 3a slowly isomerized to cis- 1-propenyl (pheny1thio)methyl ether without the formation of 2a when treated with t-BuOK (1.25 equiv) in refluxing THF.8 Thus, under these reaction conditions allylic carbanion 4, if formed, cannot be rearranged to give 2a. Not only other allylic alkoxides but also propargylic, benzylic, and even simple alkyl alkoxides undergo C-H insertion reaction to give &phenylthio alcohols 2a-m besides forming the corresponding ether 3a-m (eq 3). These results are summarized in Table I. R1R2CHOM + ClCH’SPh
t-BuOK
THF
1a-m R’R2CCH2SPh + R1R2CHOCHzSPh (3) OH 2a-m
3a-m
a, R’ = CH,=CH-; R’ = H
b, R’ = frunsCH,CH=CH-;
R’ = H
c, R’ = trans-PhCH=CH-; R2 = H d, R’ = CH,=C(CH,)-; R2 = H e, R’ = (CH,),C=CH-; R’ = H f, R’ = CH,=CH-; R’ = CH, g, R’ = H C S - ; R Z = H
h,R’=Ph-;RZ=H i, R’ = E t ; R2= H j , R’ = CH,; R2 = CH,
While a small amount of olefin addition product 5 (13%) was obtained in addition to the insertion product 2e in the reaction (2) Seyferth, D.; Mai, V. A.; Gordon, M. E. J . Org. Chem. 1970,35, 1993. (3) (a) Evans, D. A.; Baillargeon, D. J. Tetrahedron Lett. 1978, 3315 and 3319, and references cited therein. (b) Steigerwald, M. L.; Goddard, W. A,, 111; Evans, D. A. J . Am. Chem. SOC.1979, 101, 1994. (4) (a) For organosulfur carbenes and carbenoids see, for example: Block, E. “Reaction of Organosulfur Compounds”; Academic Press: New York, 1978. (b) SchBllkopf, U.; Lehmann, G. J.; Paust, J.; Hartl, H.-D. Chem. Ber. 1964, 97,1527. (c) Boche, G.; Schneider, D. R. Tetrahedron Lett. 1975,4241. ( 5 ) Under the reaction conditions of the Williamson synthesis, observations of complex reactions have sometimes been reported. For example, Corey et al. have noted a complex reaction between iodomethyl methylsulfide and secondary or tertiary sodium alkoxides.6 Carig et al. have detected the C-H insertion product of vinylidenecarbene when they treated 1-bromoalk- 1-ynes or 1-bromo 1,2-dienes with sodium alkoxides.’ (6) Corey, E. J.; Bock, M. G. Tetrahedron Lett. 1975, 3269. (7) (a) Craig, J. C.; Beard, C. D. Chem. Commun. 1971,693. (d) J . Am. Chem. SOC.1974, 96, 7950. (8) Price, C. C. Snyder, W. H. J . Am. Chem. SOC.1961, 83, 1773
0002-7863/81/1503-5965$01.25/00 1981 American Chemical Society
5966
Communications to the Editor
J . Am. Chem. SOC.,Vol. 103, No. 19, 1981
Table I. Reaction of Alkoxides with (Phenvlthiokarbene" productb (yield, %)' entry
1 2e
3 4
alkoxides (la)
p..,~~a
W
N
a
Ph-No
(1b) (IC)
(amount, equiv)
time, h
(2.5)
0.5
2ad (29)
3ad (64)
(2.5) (2.5) (2.51
0.5 0.5 0.5
2a (26) 2bd (43) 2cd (45)
3a (30) 3bd (38) 3cd (43) 3d (40) 3e (34)
C-H insertion
5 6
(2.5)
0.5
(2.5)
0.5
7
(1.25)
0.5
2d (42) 2e (38) Sf (13) 2f (41)
(2.5)
0.5
2g (22)
8
H M C H , O N a (lg)
9 10 11
PhCH,OLih ( l h ) n-PrONa (li) i-PrOLih (1j)
12
~
13
t - B u e o N .
-
E
~
~
-
(1m)
~
3f (23) 3g (22)
68 (39) 3hd (22)
(1.25)
1.o 0.5 1.o
2hd (48) 2i (9) 2j (12)
(1.25)
0.5
2kd (6)
3kd (27)
(1.25)
0.5
2md (22)
3m (18)
(1.25) (1.25) -( l ko ) N
C-M insertion
3i (91) 3j (36)
~
" Unless otherwise noted, sodium alkoxides (1.25-2.5equiv) prepared in situ from the corresponding alcohols with NaH in T H F were allowed to react with CICH,SPh (1 equiv) and t-BuOK (equivalent to alkoxides) for the period indicated. All products showed satisfactory Yields refer t o the isolated yield based o n ClCH,SPh. spectral data (IR, 'H NMR, and mass). Satisfactory analytical results were obtained for these compounds. e t-BuONa (2.5 equiv) was employed in place of t-BuOK. 5 ; [ 3,3-dimethyl-2-(phenylthio)cyclopropyl]methanol (mixture of isomers). g 6; CH,=C=CHOCH,SPh. Lithium alkoxide prepared in situ from the corresponding alcohol with BuLi was used.
of sodium prenyl alkoxide (entry 6 ) , no addition product was detected in other series of allylic alkoxides. These results present a remarkable contrast to the reported addition of dihalocarbene to allylic ethers and alcohol^.^ Seyferth et al. have reported that dichlorocarbene generated from PhHgCC12Br reacts with allyl ethyl ether to give both the addition and the C-H insertion product ~ addition of in 82% and 14% yield, r e s p e c t i ~ e l y . ~Exclusive dihalocarbenes generated under phase-transfer conditions to allylic alcohols, such as methallyl alcohol and prenyl alcohol, was also reported."-c To ascertain that the present characteristic reactivity of the a-C-H bond of alkoxides is not due to the reactivity of (pheny1thio)carbene but to an oxyanionic substituent effect, the following experiments were performed. (Phenylthio)carbene, generated under phase-transfer conditions (CH2C12-aqueous 50% NaOH using BzNEt3+Cl- as a catalyst)," was allowed to react with prenyl alcohol to give olefin adduct 5 (3.8%) together with prenyl (pheny1thio)methylether 3e (41%).1° The insertion product 2e could not be detected after thorough examination of the reaction mixture by means of column chromatography. Moreover, the reaction of methallyl methyl ether with (pheny1thio)carbene under similar conditions only gave an olefin adduct, 2-methyl-2(methoxymet h yl) - 1 - (phenylt hio)cyclopropane (36%). The present oxyanionic substituent effect which facilitates the C-H insertion seems to be based on (1) the decrease in the C-H bond energy of a l k o x i d e ~and ~ ~(2) the stabilization of the polar transition state 7 by electron-donating substitution." The ste(9) (a) Seyferth, D.; Burlitch, J. M.; Minasz, R. J.; Mui, J. Y.-p.; Simmons, H. D., Jr.; Treiber, A. J. H.; Dowd, S. R. J. Am. Chem. SOC.1965,87, 4259. (b) Seyferth, D.; Mai, V. A. Ibid. 1970, 92, 7412 and references cited therein. (c) Hirano, S.; Hiyama, T.; Nozaki, H. Tetrahedron 1976,32,2381. (d) Hiyama, T.; Tsukamoto, M.; Nozaki, H. J. Am. Chem. SOC.1976.96, 3713. (e) Kleveland, K.; Skattebol, L.; Sydnes, L. K. Acta Chem. Scand. Ser. B 1977,831,463. (10) 3e might be formed either by 0 - H insertion of (pheny1thio)carbene or by SN2 displacement reaction. (1 1) The insertion may occur by attack either on the electrons of the C-H bond (7)12 or on the hydrogen atom via the transition state with a linearly arranged PhSCH-.H-.C(O-) conformation.l' (12) (a)Doering, W. von E.; Prinzbach, H. Tetrahedron 1959,6, 24. (b) Skell, P. S.; Woodworth, R. C. J . Am. Chem. SOC.1956, 78, 4496. (c) Reference 2. (d) Gutsche, C. D.; Bachman, G. L.; Udell, W.; Bauerlein, S. J. Am. Chem. SOC.1971, 93, 5172.
reospecificity of the reaction as demonstrated below supports the validity of this concerted mechanism of the C-H insertion. Thus, the sodium alkoxides of trans- and cis-4-tert-butylcyclohexanol in the reaction with (pheny1thio)carbene ~tereospecifically'~ gave the axial (2k) and equatorial C-H insertion product (2m),respectively (entry 12 and 13, eq 4 and 5).15 As shown in Table FHzSPh lk
t
2k
-0CHZSPh
(4) 3k
OH
1m
+
CICHzSPh 2m
?CH,SPh
(5)
3m (13) (a) Benson, S. W.; DeMore, W. B. Adu. Photochem. 1964, 219. (b) Dobson, R. C.; Hayes, D. M.; Hoffman, R. J . Am. Chem. Soc. 1971,93,6188. (c) Bcdor, N.; Dewar, M. J. S.; Wasson, J. S. Ibid. 1972, 94, 9095. (14) In each reaction, the other stereoisomer could not be detected by VPC analysis of the reaction mixtures. (15) Structures of isomeric alcohols 2k and 2m are determined after the desulfurization of one isomer 2m to the known cis-4-tert-butyl-1-methylcyclohexanol (mp 69-70 OC, lit. 69-70 "C,I6 70.5-71 "C") using Raney nickel in refluxing ethanol (94%yield). Structures of 3k and 3m were determined by means of IH NMR spectral analysis of their C(l) protons: the axial proton of 3k resonates as a broad multiplet (ca.40 Hz) centered at 6 3.55, while the equatorial proton of 3m resonates as a multiplet (ca. 16 Hz) at 6 4.00.18 (16) Corey, E. J.; Chaykovsky, M. J. Am. Chem. Soc. 1965, 87, 1353. (17) Depuy, C. H.; King, R. W. J . Am. Chem. SOC.1961, 83, 2743. (18) Jackman, L. M.; Sternhell, S. "Application of Nuclear Magnetic Resonance in Organic Chemistry, 2nd ed.; Pergamon Press: 1969; Chapter 4-1.
J . Am. Chem. SOC.1981, 103, 5967-5968
I the ratio of the C-H insertion to the formal C-Metal insertion
5967
3
rises with the increase of electron-donating substitution at the olefin
8
or a-carbon site, and this validates the second factor for facilitating the C-H insertion (vide supra). R‘ R‘
//
H;--,C-OM *,-,’ a+
A
H
SPh
7
As to the formation of ether 3, there are two possible mechanisms: (1) SN2attack of alkoxide anion on ClCH2SPh and (2) the insertion of (pheny1thio)carbene into the oxygen-metal bond and subsequent protonation. In the absence of t-BuOK, sodium allylalkoxide reacts with ClCH2SPh very slowly at 0 OC, and only a trace amount of 3a is produced together with the almost quantitative recovery of ClCH2SPh after reacting for 1 h. While this result suggests mechanism 2, there still remains at present a possibility of mechanism 1 where the addition of t-BuOK provides the corresponding potassium alkoxide as an active nucleophile. A reaction using t-BuONa in place of t-BuOK was performed, giving rise to the formation of 2a and 3a in 26% and 30%yield, respectively (entry 2, Table I). Thus, at least in this reaction, it is apparent that the formation of 3a proceeds via the insertion of (pheny1thio)carbene into the oxygen-sodium bond. Work is in progress to disclose more precisely the nature and the synthetic potential of the present reaction.
Oxotrimolybdenum(1V) Alkoxides, Mo3(CL3-o)(CL3-oR)(C12-OR)3(OR)6(R = CH(CH3)2 and
CH2C(CH3)3). Synthetic Considerations
M. H. Chisholm,* K. Folting, J. C. Huffman, and C. C. Kirkpatrick
Figure 1. ORTEP view of the central Mo30(OC),,, skeleton of the Mo3O(ONe)lo molecule. All atoms are assigned arbitrary thermal parameters. Each molybdenum atom is in a distorted octahedral environment with respect to six directly bonded oxygen atoms. Pertinent bond distances in A (averaged) are Mo-Mo = 2.529 (9), Mo-psO(oxo) = 2.03 (3), M+g,OR = 2.21 (3), M+g20R = 2.02 (3), Mo-OR(terminal) trans to O(4) = 1.94 (2), Mo-OR(terminal) trans to p3-OR = 1.85 (3).
N e (Ne = CH2CMe3). Indeed, from reactions involving Mo2(ONe), and 02,we were able to isolate the green crystalline compound M 0 ~ ( 0 ) ( 0 N e ) which ~ ~ , was fully characterized by an X-ray study! An ORTEP view of the central Mo3O(OC),o skeleton of the molecule is shown in Figure 1 along with some pertinent bond distances. A retrosynthetic analysis suggested that the triangulo-Mo3(p3-O)unit could be constructed by the addition of an oxomolybdenum (6+) unit across the M e M o bond as in reaction 1. This bears analogy with the approach to cluster synthesis adopted by Stone and co-workers involving M==CR2 and M = C R
group^.^
Department of Chemistry and Molecular Structure Center Indiana University Bloomington, Indiana 47405 Received May 18, 1981
-
MO3(p3-O)(OR)10 M02(OR)6 + MoO(OR)4 (1) The oxomolybdenum(6+) alkoxides, MoO(OR)~,which were unknown, were viewed as the products of a simple replacement of a Mo=Mo bond6 by two Mo=O bonds in the oxygenolysis reaction (2). Previously it had been shown’ that the three electron 2MoO(OR)4 M02(0R)8 + 02 (2)
Recently a large class of triangulo-M3 complexes of molybdenum and tungsten have been discovered.’ Some of these have ligand, NO, readily cleaves the M-Mo bond to give two Moinvolved 8-4 cluster electrons, though most commonly the number N O bonds, which may be formally viewed as M=N-O.’ is 6. As yet little is known about the reaction pathways leading This synthetic strategy was successful: M O ~ ( O Rand ) ~ moto these compounds, though they do appear to form very readily lecular oxygen react to yield M o O ( O R ) ~compounds which, in with a variety of ligands. For example, Bino, Cotton, and Dori turn, react with M O ~ ( O Rcompounds )~ to yield M O ~ O ( O R ) , ~ recently reported2 their characterization of six compounds of compounds in essentially quantitative reactions.8 formula [ M o ~ ( ~ ~ - X ) ( ~ ~ - Y ) ( O A C ) ~ ( where H ~ ~ )X~=] ~ - . ~ H ~ ~ , Y = 0, X = Y = CCH3, and X = 0, Y = CCH3. All of these (4) Crystals suitable for X-ray work were grown from methylene chloride. were formed in the reaction between Mo(CO),, AcOH, and Crystal data, collected at -161 OC, using Mo K a radiation, gave the following: ( A C O ) ~which ~ , also yields the well-known dinuclear compound a = 35.56 (2), b = 18.97 (I), c = 19.34(1) A; space group Pbcn; Z = 8 ; d a = 1.295 g c d . Of the 10379 reflections measured in the range 6 5 28 5 M o ~ ( O A C(MLM). )~ We wish here to report our discovery of 40°, 6102 were unique. Only 55% of the unique data were observed using oxotrimolybdenum(1V) alkoxides of formula M O J ( ~ ~ - O ) ( ~ J - the criteria F 2 2.33u(F). In retrospect, this was proven to be caused by loss OR)(p2-OR)3(OR)6, where R = isopropyl (i-Pr) and neopentyl of a solvent molecule (CH2C12)and disorder of one of the ONe ligands [0(5) in Figure 11 and high thermal motions of all the neopentyl groups. The (Ne), and furthermore to show that these compounds can be structure was readily solved by direct methods and Fourier techniques and prepared in high yields by rational syntheses based on known refined by full-matrix least squares. Only the largest 1500 reflections were reactivity patterns of metal-metal bonds. used in the refinement. Molybdenum atoms were assigned anisotropic thermal During the course of studies of the reactions between M O ~ ( O R ) ~ parameters, and all other atoms were assigned isotropic parameters. Two peaks that occurred in a void in the crystal were assigned as chlorine atoms, compounds and molecular oxygen, which lead ultimately to whose occupancy refined to 0.33. The carbon of the CH2C12 molecule was M O O ~ ( O Rcompounds )~ with cleavage of the M m M o bond,’ we not discernible. Final residuals are R(F) = 0.0752 and Rw(F) = 0.0759. noted the formation of green intermediates when R = i-Pr and ( 5 ) Ashworth, T. V.;Chetcuti, M. J.; Farrugia, L. J.; Howard, J. A. K.; -+
(1) Muller, A.; Jostes, R.;Cotton, F. A. Angew. Chem., Inr. Ed. Engl. 1980, 19, 875.
(2) Bino, A.; Cotton, F. A.; Dori, Z . J. Am. Chem. SOC.1981, 103, 243. (3) Chisholm, M. H.; Folting, K.; Huffman, J. C.; Kirkpatrick, C. C.; Ratermann, A. L. J . Am. Chem. SOC.1981, 103, 1305.
0002-7863/81/1503-5967$01.25/0 . , I
,
~
~
Jeffrey, J. C.; Mills, R.; Pain, G. N.; Stone, F. G. A.; Woodward, P. In “Reactivity of Metal-Metal Bonds”; Chisholm, M. H. Ed.; American Chemical Society: Washington, DC, 1981; ACS Symp. Ser. No. 155, Chapter 15. (6) Chisholm, M. H.; Cotton, F. A,; Extine, M. W.; Reichert, W. W. Inorg. Chem. 1918, 17, 2944. (7) Chisholm, M. H.; Cotton, F. A,; Extine, M. W.; Kelly, R. L.J . Am. Chem. SOC.1978, 100, 3354.
0 1981 American Chemical Society
